MCP App Store
by KeenEthics
PostCo icon

PostCo MCP Server for Claude and ChatGPT

By PostCo

Overview

PostCo is a commerce SaaS platform and the revenue and retention engine behind ambitious Shopify brands. Upsell, Checkout+, Order Editing, and Returns work together on one shared intelligence layer to grow order value, reduce support work, and retain more revenue across the customer journey. PostCo MCP extends that platform into agentic commerce by giving retailers and their AI agents secure, read-only access to data in their PostCo Return Centre account. Retailers can ask questions in natural language, analyze returns performance, inspect operational records, and use trusted PostCo data in agentic automation workflows. The MCP can surface trends across return volumes, statuses, reasons, products, SKUs, customers, countries, couriers, refunds, exchanges, store credit, fraud, shipments, upsell revenue, and more. Retailers can move from high-level analysis to operational detail by finding return orders—including returns associated with a customer email—and retrieving the details of a selected return order. Each connection is scoped to the authenticated retailer’s PostCo account, helping agents understand and act on return data without modifying records in PostCo.

How to connect

  1. Claude: Open the PostCo listing in the Claude connectors directory and choose Connect. Claude listing PostCo exposes 9 tools here.
  2. ChatGPT: Open the PostCo listing in the ChatGPT app directory and choose Connect. ChatGPT listing PostCo exposes 9 tools here.

Tools

breakdown_metric

ChatGPT
Returns one metric for a time range split by a single dimension, ranked from largest to smallest. Reports how many groups were returned, whether lower-ranked groups were omitted, and whether any groups were withheld. Money metrics accept the dimension shop_currency, which is the only call that can answer a range spanning a change of shop currency: each group is in its own currency, so no total is implied and meta.currency is null.

compare_periods

ChatGPT
Returns one metric for two time ranges side by side, with the absolute and relative difference between them. Where a period is incomplete or the two are of unequal length, that is reported alongside the figures.

describe_metric

ChatGPT
Returns the full definition of one metric: what it counts or sums, what it deliberately excludes, which timestamp it is measured against, its unit, the dimensions and filters it accepts, and the version of that definition.

get_analytics_context

ChatGPT
Returns which retailer this connection answers for — its store domain and name — and that retailer's analytics configuration: time zone, currency and unit conventions, the date of their earliest return, the range of dates that can contain data, and the current metric definition version. Also returns the metric keys this API serves. The retailer is fixed by the connection and cannot be chosen.

get_return_order

ChatGPT
Returns one return order in full: the items being refunded and exchanged with their reasons, resolution methods and amounts, the shipping method, and the shipment with its addresses and whether its fee is customer-borne and settled. Also returns a link to this return order in the retailer's dashboard, which is where it can be reviewed, received or resolved. Narrative the customer wrote — item notes and free-text answers — is never returned; the addresses above are not narrative. Where a field has been withheld it is returned as null and named in excluded_fields.

list_metrics

ChatGPT
Lists the metric keys this API serves, each with a one-line summary, the grains it supports, the dimensions it can be split by, and whether its figures are live or periodically refreshed. The keys returned here are the only valid values for the metric argument of the other tools.

list_records

ChatGPT
Returns individual records for one of the row-level datasets over a period — fraud reports, shipment labels, question answers, upsell revenue, instant exchanges, all return items, bonus credits, or return amount adjustments. Rows may contain customer contact details and order identifiers. Narrative the customer wrote — item notes and free-text answers — is never returned; addresses and contact details are not narrative and are. Where a field has been withheld it is returned as null and named in excluded_fields. Rows are found by period, never by return order; a single return order is what list_return_orders finds.

list_return_orders

ChatGPT
Finds return orders by id, order name, status, customer email, shipping method or submission date, newest first. At least one of ids, order_names, statuses, customer_emails, shipping_method_types, submitted_at_min or submitted_at_max must be given. Returns one row per return order — its order, its customer's contact details, the shipping method the shopper picked, and the timestamps of each step it has been through — without the items on it, which get_return_order returns. Takes no period: a return order is looked up by identity, not summed over a range. Pages by opaque cursor.

query_metric

ChatGPT
Returns one metric for a time range, either as a single total or as a series by day, week or month. Each value is returned twice: as an integer for arithmetic, and as a formatted string in the retailer's currency for display.

breakdown_metric

Claude

compare_periods

Claude

describe_metric

Claude

get_analytics_context

Claude

get_return_order

Claude

list_metrics

Claude

list_records

Claude

list_return_orders

Claude

query_metric

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/postco"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/postco)